poi读取word中的目录大纲,导入

在做一个导入word的业务时,发现poi并不能很好的读取到文档中的目录和级别,这段代码做个记录

cpp 复制代码
 public ResponseResult importDirectory(MultipartFile file){
        try {
            InputStream inputStream = file.getInputStream();
            XWPFDocument xdoc = new XWPFDocument(inputStream);
            List<XWPFParagraph> paragraphs = xdoc.getParagraphs();
            List<ReadDto> readDtos = new ArrayList<>();
            for (XWPFParagraph paragraph : paragraphs) {
                String text = paragraph.getText();
                String titleLvl = getTitleLvl(xdoc, paragraph);
                if (StringUtil.isNotEmpty(titleLvl)) {
                    int level = Integer.valueOf(titleLvl);
//                System.out.println("text: " + text + ", titleLvl: " + titleLvl);
                    ReadDto readDto = new ReadDto();
                    readDto.setText(text);
                    readDto.setTitleLevel(level);
                    readDtos.add(readDto);
                }
            }
            int zeroCount = 0;//0出现的次数
            int oneCount = 0;//1出现的次数
            int twoCount = 0;//2出现的次数
            int threeCount = 0;//3出现的次数
            int curPoint = 0;//当前指针值
            for (int i = 0; i < readDtos.size(); i++) {
                int curLevel = readDtos.get(i).getTitleLevel();
                if (curLevel > 4) {
                    throw new BusinessException("暂不支持目录层级超过4层!!!");
                }
                if (curPoint == 0) {
                    zeroCount++;
                    curPoint = 1;
                    readDtos.get(i).setOne(zeroCount);
                } else if (curPoint == 1) {
                    if (curLevel == 0) {
                        zeroCount++;
                        oneCount = 0;
                        twoCount = 0;
                        threeCount = 0;
                        curPoint = 1;
                        readDtos.get(i).setOne(zeroCount);
                    }
                    if (curLevel == 1) {
                        curPoint++;
                        oneCount++;
                        readDtos.get(i).setOne(zeroCount);
                        readDtos.get(i).setTwo(oneCount);
                    }
                } else if (curPoint == 2) {
                    if (curLevel == 0) {
                        zeroCount++;
                        oneCount = 0;
                        twoCount = 0;
                        threeCount = 0;
                        curPoint = 1;
                        readDtos.get(i).setOne(zeroCount);
                    } else if (curLevel == 1) {
                        oneCount++;
                        twoCount = 0;
                        curPoint = 2;
                        readDtos.get(i).setOne(zeroCount);
                        readDtos.get(i).setTwo(oneCount);
                    } else if (curLevel == 2) {
                        curPoint = 3;
                        twoCount++;
                        threeCount = 0;
                        readDtos.get(i).setOne(zeroCount);
                        readDtos.get(i).setTwo(oneCount);
                        readDtos.get(i).setThr(twoCount);
                    }
                } else if (curPoint == 3) {
                    if (curLevel == 0) {
                        zeroCount++;
                        oneCount = 0;
                        twoCount = 0;
                        threeCount = 0;
                        curPoint = 1;
                        readDtos.get(i).setOne(zeroCount);
                    } else if (curLevel == 1) {
                        oneCount++;
                        curPoint = 2;
                        twoCount = 0;
                        readDtos.get(i).setOne(zeroCount);
                        readDtos.get(i).setTwo(oneCount);
                    } else if (curLevel == 2) {
                        curPoint = 3;
                        twoCount++;
                        threeCount = 0;
                        readDtos.get(i).setOne(zeroCount);
                        readDtos.get(i).setTwo(oneCount);
                        readDtos.get(i).setThr(twoCount);
                    } else if (curLevel == 3) {
                        threeCount++;
                        if (i < readDtos.size() - 1) {
                            int nextLevel = readDtos.get(i + 1).getTitleLevel();
                            if (nextLevel > 3) {
                                throw new BusinessException("暂不支持目录层级超过4层!!!");
                            }
                            if (nextLevel == 3) {
                                curPoint = 3;
                            } else if (nextLevel < 3) {
                                curPoint = nextLevel + 1;
                            }
                        }
                        readDtos.get(i).setOne(zeroCount);
                        readDtos.get(i).setTwo(oneCount);
                        readDtos.get(i).setThr(twoCount);
                        readDtos.get(i).setFou(threeCount);
                    }
                }
            }
            System.out.println(readDtos);//读取数据结果集合
            return status ? ResponseResult.success("操作成功") : ResponseResult.failure("操作失败");

        } catch (IOException e) {
            e.printStackTrace();
            throw new RuntimeException(e);
        }
//        return ResponseResult.failure("操作失败");
    }

    private static String getTitleLvl(XWPFDocument doc, XWPFParagraph para) {
        String titleLvl = "";
        try {
            //判断该段落是否设置了大纲级别
            if (para.getCTP().getPPr().getOutlineLvl() != null) {
                return String.valueOf(para.getCTP().getPPr().getOutlineLvl().getVal());
            }
        } catch (Exception e) {
        }
        try {
            //判断该段落的样式是否设置了大纲级别
            if (doc.getStyles().getStyle(para.getStyle()).getCTStyle().getPPr().getOutlineLvl() != null) {
                return String.valueOf(doc.getStyles().getStyle(para.getStyle()).getCTStyle().getPPr().getOutlineLvl().getVal());
            }
        } catch (Exception e) {
        }
        try {
            //判断该段落的样式的基础样式是否设置了大纲级别
            if (doc.getStyles().getStyle(doc.getStyles().getStyle(para.getStyle()).getCTStyle().getBasedOn().getVal())
                    .getCTStyle().getPPr().getOutlineLvl() != null) {
                String styleName = doc.getStyles().getStyle(para.getStyle()).getCTStyle().getBasedOn().getVal();
                return String.valueOf(doc.getStyles().getStyle(styleName).getCTStyle().getPPr().getOutlineLvl().getVal());
            }
        } catch (Exception e) {

        }
        try {
            if (para.getStyleID() != null) {
                return para.getStyleID();
            }
        } catch (Exception e) {

        }

        return titleLvl;
    }
相关推荐
2601_9659129115 小时前
PDF转Word与拆分组合操作实测:从格式保真到批量处理的方案对比
pdf·word
DS随心转APP17 小时前
生成word文档的DeepSeek底层解码与“AI导出鸭”工程化方案:一份技术架构师的全维度测评
人工智能·ai·chatgpt·word·deepseek·ai导出鸭
DS随心转插件18 小时前
ChatGPT星号怎么去除?AI 导出鸭从底层根治标记符号顽疾
人工智能·ai·chatgpt·word·deepseek·ai导出鸭
wujian83111 天前
怎么用千问生成word文档:从「格式崩」到「一键过」,AI导出鸭打通最后半厘米
人工智能·ai·c#·word·豆包·deepseek·ai导出鸭
Am-Chestnuts1 天前
豆包表格复制到 Word 后错列,先保留 Markdown 再转换
word
2601_965913002 天前
多份PDF批量转Word合并,三次踩坑后的方案对比与性能评估
pdf·word
DS随心转APP2 天前
deepseek生成的word怎么下载 AI导出鸭全平台方案技术深度测评
人工智能·ai·chatgpt·word·deepseek·ai导出鸭
Am-Chestnuts2 天前
AI 复制到 Word 图片丢失怎么办?先检查图片来源和保存方式
word
DS随心转小程序2 天前
AI 导出鸭重构转化链路,全面优化腾讯元宝输出 word 文档办公效率
人工智能·重构·aigc·word·豆包·deepseek·ai导出鸭
皮皮虾❀2 天前
千问办公+Office一站式生成+原生编辑一体化实战:从“AI生成文档格式错乱”到“PPT/Word/Excel直接出成品可商用”的交付型AI落地路径
人工智能·word·powerpoint